Facebook, measuring 'viewing' and the quest for real attention

Before we answer that question, it makes sense to look at what “viewing” is and what it should be.

My wife watches Game of Thrones. I don’t. And while I’m often in the room with her when it’s playing on the TV, I’m not paying attention. In the real world I haven’t viewed GOT. But in the advertising world my viewing duration for GOT is pretty high.

This is because the common sense definition of viewing is not the same as the technical definition. When people say they’ve watched something, they mean that it played in front of them and that they looked and listened. Therefore, common sense viewing can be defined as active attention.

In the digital world, attention is a very hard metric to measure, so “viewing” is measured as the opportunity for an advertisement to be seen. Instead of measuring attention, we only measure that the video played. Typical metrics such as impressions, views, viewability and yes, even completions, only represent an opportunity to be seen because of forced views and poor user experiences.

The problem for advertisers is that advertising only works when people pay attention to it, not when they ignore it. If advertisers can identify attention, then they can find more of it and make their advertising more efficient and more effective.

Even though attention is hard to measure with certainty, it is possible to set up conditions to ensure the highest degree of likelihood that “viewing” equals real attention. Here are three things that advertisers can do to harness and gauge real attention.

Focus on user experience

Is the video front and center with sound on? Is the ad easy to use and navigate? Does it behave as expected? Your video isn’t jumping up without the viewer’s permission, is it? Advertisers should embrace non-intrusive user experiences, which the IAB’s LEAN ads initiative brings attention to. And advertisers should look for vendors who use A/B testing rather than gut feel to ensure that ad designs are enjoyable and effective.

Make viewing optional

Being forced to watch a commercial that isn’t relevant or that they’ve already seen a zillion times is annoying for viewers. If viewers can’t block or skip your ad, they’ll mentally tune it out. By allowing people to skip your video, it’s likely that those who didn’t skip were paying attention. This makes your viewing metrics more meaningful and useful.

Identify true opt-in viewing

The entire viewer experience should be considered when defining the point at which a person’s active attention has truly been captured. Are viewers actively starting the video or turning on the sound? How easy and clear is it for a viewer to skip or close an ad? How much of the screen does an autoplaying video occupy? If it’s not the majority of the screen, then short view durations and low interaction rates may indicate that the video is simply playing on the screen without attentive eyeballs. Give extra scrutiny to incentivized views where people are often more interested in the incentive than the video. Advertisers need to work with each vendor to ensure that attention is adequately qualified and that attentive viewers are correctly identified.

Once advertisers are confident they can identify active attention, they should then ensure that their metrics represent active attention. This brings us back to Facebook.

Facebook recognises that lots of people ignore autoplay videos. So, smartly, they don’t count a video view until the video has played for three seconds. This way, views represent an opt-in (either actively by clicking or passively by not stopping). If you don’t think three seconds is enough, that’s fine because Facebook also offers 10-second views. Thus, advertisers have two opt-in points to choose from. However, Facebook doesn’t have performance metrics that correlate to these two points.

Facebook received negative attention recently because they combined two statistics that were incompatible. To compute average view duration, they took total video play time for anyone who started the video and divided by the number of people who viewed for at least three seconds.

Facebook corrected this, but their new metric doesn’t offer much value to advertisers. They now take total video play time and, correctly, divide by the number of people who started the video. With their new metric, Facebook is averaging view time for all of the people who quickly scroll past an autoplaying video (while it plays for a couple of seconds) with the people who opt-in to actually watch.

Questions that should be top of mind for advertisers are: How long did the viewers who opted-in to my video watch for and what actions did they take? How do groups of viewers from one opt-in point compare to groups of viewers from another? On Facebook, for example, this would mean comparing watch time, unique viewers, audience and engagement between 3-second viewers and 10-second viewers. How can I better understand these attentive viewers so that I can unearth more of them?

What’s clear is that advertisers should seek out solutions that not only offer great viewing experiences, but also offer metrics that capture active attention and not just playback. This will help advertisers understand how much real value is being delivered and how to find more of it.

*Insights and statistics collected from US, UK and South African data sources.

About Exponential
Exponential Interactive delivers innovative advertising experiences that transform the way audiences interact with brands across desktop and mobile. Exponential's platform fuses one of the largest global digital media footprints and proprietary data with user-centric ad formats designed to drive engagement and performance. Creativity, data and audience insights form the foundation for building smart and relevant brand engagement and brand performance solutions for advertisers and publishers. Exponential was founded in 2001 and has locations in 22 countries.

For more information, please visit www.exponential.com.